Abstract Dual-porosity and dual-permeability naturally fractured reservoir models assume that the fractures in the reservoir are connected with each other and distributed uniformly. However, in some cases, the reservoir characteristics exhibits discrete fracture systems, which means that the fractures might be unconnected and their distribution is not uniform. In this work, a new computational model is developed to compute the transient pressure behavior for reservoirs with discrete fracture system. This computational model is based on Laplace transform. The fluid flow in the fracture system and reservoir are computed separately and flux and pressure equivalent conditions in Laplace space are applied in the fracture wall to couple the fluid flow in both systems. The results suggest that the pressure response in a reservoir with a discrete fracture system has three flow regions: fluid flow nearby the wellbore, fracture-dominated fluid flow and fluid flow beyond the fracture. The fracture orientation (i.e. the distance between the fracture and the well), fracture parameters (fracture conductivity and non-Darcy effects) and fracture distribution are the main factors affecting the pressure response. In some particular situations the fracture-dominated fluid flow region in the pressure derivative curve may present two villages, which has been met in some field cases. The model provides with a tool for identifying the fracture pattern in a specific reservoir. Also, this model can be applied for optimization design of tight gas reservoir development. Introduction Classically, the fractured reservoir is modeled with Dual porosity model (1) or dual-permeability model. Those models assume that the fractures are connected with each other and distributed uniformly. The dual-porosity model also assumes that the fluid is produced from the fractures which are intersected with the well. However, in some cases, the reservoir characteristics exhibit discrete fracture system, which means that the fractures may be unconnected and their distribution is not uniform. Such kind of reservoir system was illustrated by Gao et al. (2) , as shown in Figure 1. Gao et al. (2) , also pointed out that the chance for a vertical well to intersect a discrete natural fracture is extremely small, since natural fractures in a reservoir tends to be vertical. Another type of discrete fracture system is artificial fracture system in tight gas reservoirs. To obtain economical production rate, most of wells in a tight gas reservoir are hydraulically fractured. Therefore, the whole reservoir looks like an artificial discrete fracture system. In this work, it is assumed that the discrete fracture system has the following characteristics, The fractures in the reservoir are discrete and not connected with each other; Each fracture can be described with its orientation, geometry and diffusivity; The well is not intersected with any fracture; and, 4. The fluid flow in fracture system obeys Forchheimer equation and the fluid flow in matrix system is Darcy flow. Zeng and Zhao (3) presented a model for non-Darcy flow in hydraulic fractures. For a reservoir with only one fracture existed, if the fracture is close enough to the well, then the system is similar to the system with a hydraulically fracture well.

Abstract Different gel systems have been extensively used to reduce unwanted water production and improve overall hydrocarbon productivity in many reservoirs. The main target for gel placement is to reduce channeling through fractures or highpermeability zones, while redirecting flow to less permeable unswept zones (1) . It is a common fact that many water based gels feature disproportionate permeability reduction (DPR) effect, resulting in significantly higher reduction in permeability to water compared to that of oil. This phenomenon has been widely investigated in sandstone porous media; however there is a lack of information available on gel performance in carbonate porous media. The main objective of this study is to verify the presence and extent of DPR effect when placing gels in carbonate porous medium. High molecular weight polyacrylamide and two crosslinkers, chromium(III)-acetate and chromium (III) -chloride, were used as main components of gel system. A series of brine and oil injections were performed to examine performance of gel, its stability and its contribution to DPR in carbonate porous medium. The effect of residual oil saturation was also studied by conducting experiments with and without residual oil present prior to gel placement. Gel stability study was carried out by subjecting gel to different pressure gradients under applied range of flow rates. The obtained results clearly indicate effectiveness of water based gel application to significantly reduce permeability to water in carbonate rock, as well as verifying DPR phenomenon in carbonate porous medium. Introduction Excessive water production turns into a major problem for oilfield industry when hydrocarbon fields become mature. Controlling water becomes a main objective as produced water significantly impacts economic profitability of the reservoir. There are many methods available to reduce water production problems, and conformance control methods are considered by far the most successful. Conformance control describes any technique that encourages the drive mechanism to mobilize rather than avoid unswept hydrocarbon zones (5) . Application of chemical based treatments, including polymers and gels are among the most widely used. Crosslinked polymeric gel systems have been extensively used for water shutoff treatments to reduce water channeling through fractures and highly permeable zones. The main objective of the gel placement is to diminish water production without significantly damaging oil productivity, leading to improved overall hydrocarbon recovery. The most important target for gel placement is to ensure maximum gel penetration to high permeability, watered-out zones while minimize gel penetration to oil-bearing zones (1) . Gel application is considered one of the most aggressive types of conformance control as it can totally block certain zones of porous medium (6) . Gels can be applied to both injection and production wells (3) . Among the advantages of gel application is their flexibility for pumping without a work-over rig, high control of setting time, a deeper penetration into the formation, ease of cleaning, lack of milling time, and an easy removal from the wellbore by water recirculation (4) . The most widely used gel system, used in the past years, has been based on chromium(III) crosslinker. ChromiumIII) crosslinked with polyacrylamide gels considered to be good choice materials to selectively restrict a flow to the water phase.

Abstract In the Court field, the middle Bakken Sand Pool has been operated as a heavy oil waterflood for over 15 years. However, unrecovered oil volumes in the pool remain attractive for improved recovery schemes. Recently, the Court middle Bakken reservoir model was updated to evaluate the potential for downspacing and waterflood optimization of the reservoir. The potential for 20 acre downspacing for both infill drilling and additional water injection was identified by this study. As part of the additional development program, an injectorproducer well pair was drilled to create a new injection pattern with reduced inter-well spacing. Reservoir pressures, water saturation and effective permeability to water in the pattern were determined by RFT and log data. Pattern characterization was complemented by an interference test, including all the wells in the pattern. Analytical and numerical tools were used in the test design. The most comprehensive results were obtained by local grid refinement of the pattern area in the full field simulation model. Unavoidable interference with a neighboring pattern during the test was predicted. In order to account for this effect, it was decided to run the test with the new injector active intermittently. The complete test design and analysis of the results are described. In addition, a comparison of pressure measurements by downhole gauge and acoustic well sounder are presented. Introduction The Court field, located in west central Saskatchewan, produces 17 °API heavy crude from the middle Bakken Formation, using vertical wells and a waterflood recovery scheme. The Court Bakken Sand was discovered in November 1981. The Court Bakken Sand Voluntary Unit No. 1 was formed in 1988 and waterflood commenced with only 2% of OOIP produced. There are 39 producing oil wells and 21 water injection wells currently in operation within the main pool. The current recovery factor is ~20%, with some wells producing up to 30m3/d oil. Recently, Majcher et al. (1) presented a geological and engineering overview of the reservoir. The middle Bakken Formation was deposited as NE-SW trending sand ridges. The pool has undergone significant collapse of the Torquay Formation with the subsequent creation of extensive post-depositional sinkholes within the Bakken. This structural deformation interrupts the lateral continuity of the sand ridge and its characterization is the most challenging to further field development and effective waterflood management. Another important Bakken heterogeneity is the change in well log signature that represents the variation in facies at the ridge edges (2) . This facies distribution within the middle Bakken correlates with the waterflood response. Majcher et al. (1) identified the potential for infill drilling of producers and injectors near the edges of the pool. An injector-producer pair was recently drilled in Section 36-033-28W3M on 20 acre spacing. An interference test with four wells was conducted to investigate how the heterogeneities described above affect the waterflood performance of the newly configured pattern. Test Objectives An interference test was conducted between December 2005 and January 2006 in the new pattern to investigate if the new injector 91/7-36i is hydraulically connected with the producers in the pattern.

Abstract The fundamentals of Steam Assisted Gravity Drainage (SAGD) steam chamber development are now well understood through Butler's analytical models, and extensive field and laboratory testing. However as industry continues to extend SAGD to new reservoirs and looks towards SAGD wind down at the end life of the projects, it is important that we recognize the value of not only understanding the steam chamber but also of the movement of fluid in the reservoir. The Dover SAGD Pilot is the most mature pilot of its kind in the world. A study of this Pilot has been undertaken in an attempt to understand the behavior of the fluid within and in front of the steam chamber. The economics of SAGD are significantly impacted by the cost of generating steam. At roughly 1mcf/bbl of bitumen produced for an SOR in the range of 2.3–2.5 m 3 /m 3 , natural gas is the single largest operating cost in a SAGD project. Water movement within the reservoir can impact the natural gas consumption wherein warm steam condensate not recovered must be replaced in the process by colder make-up water, decreasing the heat efficiency of the steam generation. Further, where water loss to the reservoir is high, the steam-oil ratio (SOR) may be negatively impacted. As we approach the 20 th anniversary of the initiation of the Dover Pilot, the cold water injection test performed prior to any thermal operations taking place is revisited here. Understanding the transmissibility of water in the reservoir is key to choosing the optimal operating pressures and maximizing the value of a project. It has been widely published 1,2 that the injection of non-condensable gas (NCG) into SAGD chambers will result in the accumulation of the NCG at the top of the chamber, cooling the chamber. The lower temperatures within the chamber cause the viscosity of the bitumen to increase thereby reducing the bitumen production rate. This has been suggested as a method of winding down steam chambers as they reach their economic producing limits 3,4,5 . From April 1998 to May 2002 NGC was injected with steam at the Dover Pilot. The gas volume injected at reservoir conditions was triple the volume of the produced bitumen over that time. The SAGD chambers did not behave as predicted. The bitumen production rate did not fall off any more than would be expected from a mature steam chamber and live steam was still detectable through the thermocouples within the steam chamber. Further, an increased overall recovery was observed, most likely from the gas assistance in the production of previously inaccessible reserves. The simulation model developed to describe, as well as further observations regarding the behavior of NCG in the reservoir, are discussed. Introduction Geographically located in northeast Alberta, the Athabasca Oil Sands deposit forms part of the western Canadian oil sands. With an estimated 1.7 trillion barrels of oil in place, it is arguably the single largest oil deposit in the world. SAGD, developed by Butler 6 in the early 1980's, is to date, the most successful in-situ method of exploiting this resource.

Abstract The use of semi-analytic methods for correcting flow equations to accommodate changing gas properties with pressure, has become increasingly common. It is a mainstay of modern production decline analysis as well as gas deliverability forecasting. The use of pseudo-time is one method which enables a time-based correction of gas properties, honoring the gas material balance within the time-based flow equation. By using pseudo-time, the analytical well / reservoir models, derived for the liquid case (slightly compressible fluid) can be modified for gas by re-evaluating the gas properties as the reservoir pressure depletes. These gas correction procedures are well documented in the literature. Also well documented is the iterative nature of the gas properties correction methods, as original gas-in-place is a required input into the equations. The pseudo-time correction is based on the average reservoir pressure and works very well for boundary dominatedflow. However, when transient flow prevails, the pseudo-time concept is not valid and its use can create anomalous responses. This will occur in low permeability systems or in reservoirs with irregular shapes, especially where some of the boundaries are very distant from the well. The semi-analytic gas correction has a "representative pressure" at its root, which, in the existing models, is always the average reservoir pressure. We propose a straight-forward modification to the determination of this pressure as follows. The representative pressure ought to be based on a "radius of investigation" or "region of influence" (in the case of nonradial systems), rather than the average reservoir pressure. Inthe case of a depleting system, the representative pressure would be the same as the average reservoir pressure. The following paper outlines the proposed procedure and illustrates its advantages over the existing method, by using synthetic and field data examples. Introduction Literature on the derivation and usage of pseudo-time is Prevalent (2) (3) . The definition that will be used in this paper is shown below. (Equation) (Available in full paper) The above is used in the pseudo-steady-state equation for gas, which is at the core of most modern production decline analysis methods. It is also used in analytical well / reservoir models, whose conventional formulations are only valid for slightly compressible fluids with constant properties over a given pressure range. These models enjoy widespread usage for both history matching and forecasting, and their inclusion of pseudo-time for gas reservoirs is vital. To illustrate the value of pseudo-time, let us take the simple case of a vertical well in the center of a circular gas reservoir. We will assume constant rate production and pseudo-steady state conditions. Thus, the model that describes the pressure response at the well can be reduced to the pseudo-steady-state equation for gas (5) . (Equation) (Available in full paper) The "f(t)" in equation (2) is the chosen time function. Figure 1 shows the pressure response plotted against time for two cases f(t) = time (t) and f(t) = pseudo-time (ta). Also compared is the numerical solution using the same input parameters.

Abstract Conventional horizontal well transient response models are generally based on the line source approximation of the partially penetrating vertical fracture solution1. These models have three major limitations: wellbore pressure is computed at a finite radius outside the source; it is impossible to compute wellbore pressure within the source, it is difficult to conduct a realistic comparison between horizontal well and vertical fracture productivities, because wellbore pressures are not computed at the same point, and the line source approximation may not be adequate for reservoirs with thin pay zones. This work attempts to overcome these limitations by developing a more flexible analytical solution using the solid bar approximation. A technique that permits the conversion of the pressure response of any horizontal well system into a physically equivalent vertical fracture response is also presented. A new type curve solution is developed for a horizontal well producing from a solid bar source in an infinite-acting reservoir by means of Newman's product solution2. Analysis of computed wellbore pressures reveals that error ranging from 5 to 20% was introduced by the line source assumption depending on the value of dimensionless radius (r wD ). Computations show that for r wD ?10 −4 the transient response of a horizontal well is identical to that of a partially penetrating vertical fracture system, and for r wD <0.01 the transient response of a horizontal well is indistinguishable from that of a horizontal fracture system. Type-curve plots for the ranges 0.01?dimensionless length (L D )?10, and 10 −4 ? r wD ?1.0 are presented. A dimensionless rate function (®-function) is introduced to convert the transient-response of a horizontal well into an equivalent vertical fracture response. A step-wise algorithm for the computation of ®-function is developed using Duhamel's principle. This provides an easier way of representing horizontal wells in numerical reservoir simulation without the rigor of employing complex formulations for the computation of effective well block radius. Introduction Conventional models for horizontal well test analysis were mostly developed during the 1980s. The rapid increase in the applications of horizontal-well-technology during this period led to a sudden need for the development of analytical models capable of evaluating the performance of horizontal wells. Ramey and Clonts3 developed one of the earliest analytical solutions for horizontal well test analysis based on the line source approximation of the partially penetrating vertical fracture solution. Conventional models 4–16 assume that a horizontal well may be viewed as a well producing from a line source in an infinite-acting reservoir system. These models have three major limitations: wellbore pressure is computed at a finite radius outside the source; it is impossible to compute wellbore pressure within the source, it is difficult to conduct a realistic comparison between horizontal well and vertical fracture productivities, because, wellbore pressures are not computed at the same point, the line source approximation may not be adequate for reservoirs with thin pay zones.

Abstract As our industry focuses more on natural gas production and on wringing the last drop of value from each property, it has become increasingly important to properly characterize and predict the solution gas performance from our oilfields. This paper will review historical techniques for predicting solution gas production under various common process mechanisms including depletion, weak and strong water drive, gas-cap drive, and production of a volatile oil. The paper will also discuss situations where numerical simulation may be required instead of using standard analytical techniques. The paper will present examples from the Western Canadian Sedimentary Basin that illustrate solution gas performance under each of the drive mechanisms mentioned above. Results are displayed using the "equal-value" concept, which shows the progress of each oilfield from its early life where revenue is dominated by oil sales to its later life when revenue becomes increasingly dominated by solution gas production. Introduction to the Problem The prediction of solution gas production is often taken for granted in oilfield forecasts, but is a multi-faceted problem when one considers the many components that impact the process (fluid and rock properties, interactions between fluids/rocks, geological properties, drive mechanisms, wellbore conditions, etc). Years ago, accurate predictions of solution gas volumes were less important because there was little or no value associated with the product. Today, the value of natural gas is essentially the same as oil on a heating value basis, so much more attention is paid to the prediction of solution gas volumes. Further, environmental pressures are driving government and industry to conserve all produced gas; and obviously it is important to understand how much gas will be produced to ensure installation of the appropriate conservation scheme. On its face, solution gas prediction is a deceptively easy problem: pressure declines, gas evolves from the oil, and is produced. So, if one completely understands the production mechanism and can accurately predict future pressure decline, and understands the PVT properties that govern the release of solution gas, and knows the rock properties that govern the trapping and flow of gas, and has a good picture of the overall geological model, then it can be fairly simple to predict solution gas production. Further, the interplay between these factors can result in non-unique solutions. Since the primary focus is mostly on oil volumes, appropriate attention is not always paid to the variables that govern solution gas production. The most common analytical methods for predicting solution gas production were developed by Tarner (Ref. 1) and Muskat (Ref. 2). These methods use material balance principles and a dynamic producing GOR to predict reservoir performance at pressures when the gas saturation exceeds the critical gas saturation. However, a number of simplifying assumptions were made in these analytical treatments, including thin horizontal reservoirs with negligible gravity forces (ie. no gas percolation). Other analytical approaches have been proposed that mitigate concerns with the Tarner and Muskat methods; however, all analytical methods have their own limiting assumptions.

Abstract Stimulation of carbonate formations by acid dissolution of the rock has been an efficient and successful method of bimproving production in oil and gas wells. Hydrochloric acid is the normal fluid of choice. However, in high temperature applications corrosion issues limit usage, especially in chrome completions. Acetic acid has been used with some success and with adequate corrosion protection. But due to its low reactivity at higher temperatures, the efficiency with which a gallon of acid dissolves the formation is perceived as low. This perception comes from reaction efficiency of acetic acid reported in the literature ranging in values from 90% at 25 °C to 40% at 121 °C for 2 to 15 wt%, respectively. Acetic acid reaction on calcium carbonate is controlled by its small dissociation constant, 1.754E-05 at 25 °C (77 °F) and therefore is labeled a weak acid. Abstract The effects of permeability variable in formation on pressure derivative curve are studied in this paper. We view the general heterogeneous reservoir as a base homogeneous reservoir modified with multiple variable permeabilities at some locations. The sections where permeabilities are different from the base permeability are called as the variable permeability sections. A heterogeneous reservoir with only one variable permeability section can be well defined as long as the permeability, the size and the location of the section are specified. Its pressure derivative deviation from that of the base homogeneous reservoir has been thoroughly studied. Results show that the start time and the value and the occurred time of maximum magnitude of the pressure derivative deviation suggest the start position, the permeability and the end position of the variable permeability section, respectively. In order to analyze a heterogeneous reservoir with multiple variable permeability sections, we proposed that its pressure derivative difference with respect to the base homogeneous reservoir is the summation of the pressure derivative differences, with respect to the base homogeneous reservoir, of the single-section multiple heterogeneous reservoirs, each of them possesses only one variable permeability section. This method has been proved and verified in the reservoirs with radial and areal permeability distribution using both analytical and numerical methods. Applications show that this method provides a useful clue for heterogeneity reservoir well testing analysis. If the test noise can be ignored in pressure derivative curve, this method is very practicable for well testing analysis of variable permeability reservoirs. In the cases where pressure noise makes the pressure derivative zigzag, some de-noising methods, for example Schroeter's deconvolution method, wavelets and optimal model, can be used to de-noise pressure data in order to get smooth pressure derivative curve. Then, this de-noised test data can be diagnosed using the method proposed. Introduction Traditional well testing analysis tends to determine an overall permeability, which cannot reflect the variation of permeability in formation. In our experiences on practical well testing interpretation, we often encounter the situations that we can match the shape and the trend of the pressure derivative curve perfectly, but we cannot match the slight waves in the pressure derivative curve. Generally there are two sources that produced these kinds of waves, pressure measurement noise and the response of heterogeneity of the reservoir. The waves generated from the pressure measurement noise are random and discontinuous, while those from the heterogeneity of the reservoir behave continuously and smoot

Abstract A fully coupled geomechanics-reservoir model is developed to simulate the directional changes in permeability due to fluid injection and production in a reservoir. The model is implemented numerically by fully coupling a geomechanics model with a single-phase reservoir flow model using finite element method. A strain-induced permeability model is developed based on the analysis of grain fabric of intact and sheared oil sand specimens using thin section imaging method. It can quantify the changes in permeability when material experiences shear deformation. In addition, the directions of the principal values of permeability are not restricted to some arbitrary axes, and governed by the induced strains. Thus, the effects of stress paths and stress level are implicitly considered through effective stress-strain constitutive laws. The transient pressure response of a water injection test in a horizontal well is analyzed. Parametric studies are also conducted to investigate the effects of permeability, deformability and initial stress condition on the injection pressure. It is found that the changes in permeability resulting from dilations of the oil sands cannot be captured using the conventional permeability model as a function of volumetric strain, because the volumetric strain is small. However, the strain-induced permeability model can accurately reflect the directional increase in permeability during water injection. It is shown that the induced pore pressure is relatively insensitive to the deformation modulus of the reservoir as compared to the permeability. The initial stress condition dominates the propensity of hydraulic fracturing. Introduction Geomechanics plays a key role to account for rock deformations due to pore pressure and temperature changes resulting from fluid injection and production in a thermal reservoir. During the fluid injection and production, absolute permeability at any given location may change in response to localized changes in stress within the rock pore system owing to changes in reservoir pressure. Pore pressure changes can in turn influence the effective reservoir stress (overburden pressure minus pore pressure) which can alter the pore geometry of the reservoir rock, specially the shape and dimension of the pore and the pore throats 1–7 . The geomechanical behavior of porous media has become increasingly important in deformable and fractured reservoirs. In the current geomechanicsreservoir coupled simulators, the permeability of the rock is assumed to be isotropic, transversely or orthogonally isotropic 8–9 . In other words, the direction of flow is aligned with the pressure gradient. This simplifying assumption is not valid in many cases. For example, the permeability of fractured rock is substantially dependent on microfractures, therefore, the dominating orientation of the fractures can make the direction of flow different from that of the pressure gradient 10 . Another example with permeability anisotropy is demonstrated in thermal recovery processes, such as cyclic steam stimulation and steam assisted gravity drainage. In these thermal recovery processes, shear dilation deformation takes place in local areas due to the increase of pore pressure, thereby enhancing permeability in these areas 11–12 . Clearly, an understanding of permeability anisotropy is important.

Abstract Steam-Assisted Gravity Drainage (SAGD) is an in-situ thermal recovery technique used at Petro-Canada's MacKay River Project. SAGD utilizes pairs of horizontal wells placed near the bottom of the reservoir. Steam is injected into the upper well to reduce the viscosity of the bitumen. The bottom producing well collects the produced fluids. The objective of this study is to develop an analytical well test method that would allow a measure of steam chamber volume and potentially the timing and location of steam chamber coalescence. This well test method would also prove useful in determining an analytical solution for near well bore characteristics and reservoir boundaries. A pseudo-compositional thermal simulator was used to generate reservoir pressure responses by shutting off an injector at different periods of time and monitoring the pressure falloff while continuing with production. Confined, unconfined and coalesced 2-dimensional sinksource well pair models, based upon the geology and drilling pattern at Petro-Canada's MacKay River Project, were used in the study. The effect of the magnitude of the vertical permeability and the vertical to horizontal permeability ratio on the shut-in pressure response was also studied so that this method could be applied to other reservoirs of differing geology. Pressure response type curves were generated and the relationships between pressure drop at the injector, shut in time, and steam chamber volume were determined. We show that the pressure response type curves are unique for different types of SAGD wells and levels of steam chamber development. This information will aid in determining the timing of operational conversions as well as determining recovery factors and possibly locations of reservoir boundaries as well as steam chamber coalescence. Introduction Steam Assisted Gravity Drainage (SAGD) is an in-situ thermal recovery technique that uses pairs of horizontal wells to produce bitumen. Steam is injected from the upper well into the formation to reduce the viscosity of bitumen. Gravity is the primary driving force as the steam rises and forms a steam chamber that gradually sweeps through the reservoir. The steam that is injected into the upper well creates a steam chamber that enables the heavier condensate and heated bitumen to flow downwards into the bottom producing well. (1) In a commercial SAGD project it can be expected that as the steam chambers mature and sweep through the reservoir, the well pairs will begin to interact with each other and the steam chambers will coalesce. Coalescence is defined as the point at which separate adjacent steam chambers merge to form one large steam chamber. Coalescence is used for timing operational conversions to secondary recovery schemes such as solvent injection. As a result of the planning that is required for the operational conversion, a method of measuring the volume of the steam chamber would prove useful in estimating the timing of coalescence and the subsequent operational conversions. Considering there are 25 well pairs at MacKay River, it is also necessary to know which well pairs have coalesced so as to know the well pairs in which to institute operational conversions.

Abstract Solution gas libration is one of the key factors contributing to the success in solution gas drive in heavy oil reservoirs. Unconsolidated sands dilate with decreasing effective stress, and contract with increasing effective stress. This paper describes special experiments that were designed to investigate the effect of sand matrix deformation on the bubble nucleation and growth in heavy oil, and the oil production. The experiments involved flow of heavy oil with dissolved gas in reconstituted sand packs. Gas bubbles were allowed to develop in dilated sand packs by lowering the effective confining stresses in an undrained (closed) system. Solution gas in heavy oil was allowed to produce from the sand packs in constant depletion rates. Production rates under various conditions were compared. Based on the experimental results, a mechanistic model for solution gas drive in heavy oil was proposed to explain the linear relationship between the oil production and depletion pressure. Introduction The formation of dispersed gas bubbles (or foamy oil) in the heavy oil has been postulated to be an important factor contributing to the success in primary production of heavy oil reservoirs 1, 2 . It has been hypothesized that the foamy nature of the heavy oil maintains the released solution gas dispersed in the continuous oil phase, which is very different from the convention oil behaviour. The flow behaviour, pressure responses and production rates of solution gas drive in heavy oils have been studied by several investigators using depletion tests (e.g., Sheng et al. 3 ; Wong et al. 4; Zhang et al. 5 ). The depletion tests results indicate that the recovery factors are higher in fast depletion tests of pressure decline rates of 2.1 to 3.5 kPa/min than those in slow depletion tests of pressure decline rates of 0.3 to 0.5 kPa/min. The total oil recovery under step pressure declines can be as high as up to 35 to 45% 4 . In the constant pressure decline rates, the total oil depletion lies in a range of 17 to 29% 5 . It appears that the viscosity or temperature plays a minor role in the production. These findings provide some valuable insight into the solution gas drive mechanism. However, the soution gas-oil flow properties, such as total compressibility and mobility, have not been quantified. How does the deformable reservoir sand matrix affect the solution gas drive behaviour? This paper attempts to answer some of these questions by conducting some specially designed depletion tests. TESTING MATERIAL AND EQUIPMENTS A schematic diagram of the pressure depletion test setup is illustrated in Figure 1. The test sand was packed inside a core holder of 41.2 cm in length and 3.8 cm in diameter. The core holder composed of a flexible Vitron sleeve and two steel end platens was subjected to a confining overburden pressure inside an insulated highpressure cylindrical chamber. A high-precision screw pump was connected to the high-pressure steel chamber to regulate the pumped water volume and the pressure.

Abstract Foamy oil flow is believed to be an important factor in providing higher than expected primary recovery factors in Canadian heavy oil reservoirs exploited with the cold heavy oil production (CHOP) technology. It is widely accepted that, under field conditions used in CHOP, the gas released from solution starts to flow while remaining dispersed in oil. However, our understanding of the events that occur between the nucleation of gas bubbles and the start of dispersed gas flow is incomplete. The objective of this work was to examine these events using a variety of experimental tools including an etched glass micro-model, Magnetic Resonance Imaging (MRI) and depletion tests in long sand-packs. An etched glass micro-model was used to visually examine the formation of gas bubbles when pressure is reduced and gas starts coming out of solution. These tests showed that the gas release does not always start in the part where the pressure would be expected to be lower, i.e. near the production end. A special MRI apparatus was used for in-situ measurement of the gas saturation along the length of a core, during a depletion test, in which one end of the core was open to flow, while the other end remained closed. Two comparative tests were carried out with a viscous heavy crude oil and a mineral oil of low viscosity, using the same porous medium and same pressure decline rate. A comparison of the results for these two systems showed that unlike the non-foamy system, in the foamy heavy oil test, the first bubbles appeared near the closed end and the intensity of gas release slightly decreased from closed end towards the production end. The initial pressure response of a 2-meter long sandpack to fast depletion was measured at several different locations in the sand-pack. The recorded pressures show a rapid decline in pressure followed by a substantial bounce back. Three different oils were used in this work and the results are discussed in the light of information obtained from the micromodel tests and the MRI tests. Introduction Several recent papers have noted that often the solution gas drive process operates differently in heavy oil reservoirs compared to the conventional oil systems (1–5) . This is specially true in Canadian heavy oil reservoirs when the so-called cold heavy oil production (CHOP) process, which involves production of substantial volume of sand with the oil is used. Both the oil production rate and the oil recovery factor are much higher when the sand is produced into the wells and transported to surface with the oil. Oil production rates have been reported to be more than ten times the flow rate predicted by Darcy's law and the projected recovery factors are two to three times higher than what would be predicted by the conventional solution gas drive theory (6) . It is believed that the sand production increases the fluid mobility in the near well zone by increasing the permeability in the affected zone (7) .

Abstract After a well starts flowing, a larger and larger portion of the reservoir contributes to production. At any given time, the radius of the portion of the reservoir that demonstrates a pressure gradient and contributes to production of fluid is the radius of drainage (ROD). The time required for the entire reservoir just to be able to contribute production is the stabilization time. Estimating the ROD and stabilization time is very important in well test design and production optimization. It has been a great challenge to estimate accurately the ROD and stabilization time because of inherent uncertainties with respect to the rock and fluid properties. This study examines the effects of criterion values on the estimated values of the ROD and stabilization time. As expected, estimated values of the ROD and stabilization time vary considerably, depending on the suggested criteria. The primary objective of this study is to recognize and appreciate the importance of criterion values for defining the ROD and stabilization time. Generalized correlations have been proposed that allow one to determine the ROD and stabilization time as a function of the criterion values. The relationship between a pressure criterion and the corresponding rate criterion has been examined also. Introduction The concepts of ROD and stabilization time are commonly used in reservoir engineering and in well test analysis. Estimating the ROD is very important on many counts. A well-test analysis provides important reservoir information based on the area sampled within the ROD. It is important to know the extent of the reservoir that is being sampled when determining the parameters like permeability and storage capacity from a well-test analysis. In other words, the well test analysis provides the global values of the reservoir parameters that are valid over the radius of investigation. 1 Thus, the obtained reservoir information is good for the region within the ROD. In addition, knowing the ROD helps optimize the locations of new wells to be drilled in a field. It is very difficult to identify the well-test run time without an estimate of the ROD and stabilization time. The ROD concept has both quantitative and qualitative importance in well-test design and analysis. This distance is dependent on the way the pressure response propagates through the reservoir. It is also related to formation rock and fluid properties and elapsed time. Thus, the ROD concept presents a guide for well test design. This concept can be used also to estimate the time required to test to the desired depth in the formation. However, estimating any ROD has been dependent on the assumed level of the criterion for pressure or flow rate. As a result, there can be substantial variations in the estimated magnitude of the ROD. Depending on the criterion parameters and their values, a number of definitions have been proposed for the ROD and stabilization-time equations. Daungkaew et al. 2 have provided a comprehensive account of these efforts. 3–16

Abstract A general procedure for characterizing complex reservoirs utilizing their permeability tensor is being developed by integrating data and methods from different disciplines. Permeability tensors for geologically defined fracture patterns are derived, and finally these smallscale descriptors are incorporated into a reservoir simulation program capable of handling full tensor permeabilities. The application and convenience of the method presented in this paper is illustrated with a field example from a naturally fractured reservoir. Introduction For many years substantial research has been conducted in the areas of geosciences and engineering in order to characterize naturally fractured reservoirs. Numerous approaches have been presented to properly overcome this difficult task. Geoscientists have focused their research towards understanding the process of fracturing (rock mechanics) and the subsequent description of fracture characteristics such as density and orientation. Engineers, on the other hand, have focused their attention to the description of the fluid flow in the fracture systems and in the development of accurate models (reservoir simulators), to reproduce the history and predict the hydrocarbon production for these complex systems. One of the first matrix-fracture models was presented by Warren and Root [1] , who presented an idealized sugar cube model with two classes of porosity a primary porosity that is intergranular, and a secondary porosity that is induced by fractures. Even though the sugar cube model has been widely accepted as the forerunner of the modern interpretation of dual-porosity systems, its limitations in describing the behavior of some complex fractured reservoir systems have been observed. It is now well known that almost all fracture systems are much more complicated than the suggested model of three orthogonal sets of uniform fractures. One of the most important factor that has been identified as a necessary addition to improve the overall description of such complex reservoirs is the definition of a nine-component permeability tensor for the fracture system. This tensor is used to model fluid flow in complex reservoirs with multiple zones of directional permeability, where the orientation and magnitude of the principal permeabilities may vary between different zones in the reservoir. Snow [2,3] studied the convenience of using mathematical equivalents of parallel plate openings to simulate fractures dispersed in orientation, distributed in aperture and of arbitrary spacing. Models for fractured media which contained any number of planar conductors of any orientation and any fine aperture were presented A key assumption was that all the conduits have smooth parallel plane walls of indefinite extent (infinite fractures) and an arbitrary aperture. As a result, a permeability tensor could be obtained by superposition of contributions due to the fractures and due to the permeable matrix. Long et al [4,5] addressed the more realistic scenario of finite or discrete fracture systems, where properties such as shape, orientation and location of the fractures in an impermeable matrix were considered to be random variables. A two-dimensional model was presented and later expanded to a three dimensional model, utilizing statistics to carry on the simulations of fracture systems.

Abstract Impulse tests, Perforation Inflow Diagnostic (PID) tests and Closed Chamber tests are all similar in methodology. By whatever name, they provide simple, quick, cost-effective snapshots of maximum reservoir inflow characteristics. In a matter of minutes a field Engineer can provide estimates of the maximum AOF or IPR capability, the skin effect and flow capacity of a zone. Liquid influx can also be quantified. Applications include pre-frac reservoir assessments, multiple-layer perforation inflow distribution diagnostics (PIDD), establishing initial reservoir pressures, evaluating moveable gas in shaley zones or very low permeability horizons (reserves additions), flowing or pumping oil well tests, closed chamber drillstem tests (CCDST), and surface casing vent tests. With some conventional flow &buildup tests and production history as background, PID response can be correlated to stabilized gas rates for instantaneous tie-in or frac decisions, a benefit to expedient shallow gas exploitation. For all of these tests, data has been recorded in realtime at surface. Rigorous testing utilizes modern electronic surface data acquisition equipment and various software. A proprietary hardware &software system is also available. In it's simplest, most cost-effective and exploitive form, measurements with a hand-held pressure gauge are known to be 'good enough'. Analytical theory for these tests is well documented. While the tests have been around for some time, the techniques are under-utilized in Western Canada. Introduction The primary purpose of this paper is to illustrate various applications and usefulness of impulse testing with working field examples. Over the past several years the authors have experienced growth in the number of requests for these tests and success in their application in Western Canadian Sedimentary Basin gas fields. Surging interest in impulse testing is due to technological advancements, present-day economics, modern business interests and changing strategies where exploitation has become a new Petroleum Engineering discipline. Impulse testing has been around for some time and analytical theory is well documented. A comprehensive analytical process, however, does not presently exist. Hence, a secondary purpose of this paper is to bring together the various test types and references to illustrate the practicality of developing impulse analysis software. Better analytical tools are required to fully appreciate and maximize benefits of these quick, simple cost-effective techniques. EXPANDING APPLICATIONS Over the past several years impulse testing applications have been expanding, concurrent with ever increasing gas development and primarily for by-passed pay evaluation. Technological advancements with electronic surface data acquisition equipment (surface pressure recorders) has made impulse testing more practical. Surface recorders are now becoming commonplace tools for companies to access. Advanced completion strategies and underbalanced perforating technology means that accurate and viable impulse data can be collected at surface. Introduction of coiled tubing frac's has allowed companies to specifically target very thin low permeability streaks where an impulse test may provide the only feasible pre-frac assessment tool. Lower and lower stabilized flow rates are a reality of present-day and future economics. As well, ultimate recoverable reserves are more important then ever to a companies net present value. Hence, companies are pursuing by-passed pay in very low permeability horizons.